The Center for AI Safety (CAIS) released a paper this week detailing ways humans can better understand and control artificial intelligence, including detecting lies.
An organization dedicated to the safe development of artificial intelligence released a 'breakthrough paper' it said will help humans better control the technology as it spreads. 'We can’t trust AIs if we don’t know what they are thinking or how they work on the inside,' Dan Hendrycks, director of the Center for AI Safety, told Fox News Digital.
According to the CAIS, the paper demonstrated ways humans can control and detect when AI systems are telling truths or lies, when they behave morally or immorally, whether they act with emotions such as anger, fear and joy, and how to make them less biased. The paper also looked at ways to develop systems that can resist jailbreaks, a practice where users can exploit vulnerabilities in AI systems and potentially use them outside desired protocols.
notes that modern AI systems have been notoriously difficult for humans to understand, which also makes it hard for users to understand AI decision-making. Those concerns have also been shared by Congress, with Senate Majority Leader Chuck Schumer, D-N.Y. calling AI explainability 'one of the most important and most difficult technical issues in all of AI' in remarks at the Center for Strategic & International Studies earlier this year.
We’re forming a sort of ‘internal surveillance’ for AI systems, ensuring they aren’t trying to trick us,' Hendrycks said. 'Deception in AI is a real concern, and our research is a key step towards providing tools to prevent these risks.'
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Nobel Prize in physics awarded to researchers looking at electrons in atomsThe award was announced Tuesday from Stockholm.
Read more »
Nobel Prize Given To Researchers Behind mRNA Tech In Covid VaccinesThe mRNA technology was used to rapidly develop and deploy two effective vaccines that offered protection against the Covid-19 pandemic.
Read more »
Researchers build and test a framework for achieving climate resilience across diverse fisheriesWhat makes for a successful climate-resilient fishery, one that sustainably produces resources for human benefit despite increasing climate stressors and human impacts? It's a question that faces present and future fisheries, their practitioners and fishing communities as the world turns to the ocean to feed its growing population.
Read more »
Researchers Tested AI Watermarks—and Broke All of ThemIn a new paper out this week, a group from the University of Maryland found it's too easy to evade current methods of watermarking—and even to add fake watermarks to real images.
Read more »
Researchers identify tiny, unique sea creature photographed in 2018An international team of zoologists and parasitic worm specialists has identified an odd sea creature captured by an underwater photographer in 2018 off the coast of Okinawa. In their project, reported in the open-access journal Current Biology, the group obtained and studied a sample of the sea creature.
Read more »